5 Standard NodeClasses

5.1 Overview

Clause 5 defines the NodeClasses used to define Nodes in the OPC UA AddressSpace. NodeClasses are derived from a common Base NodeClass. This NodeClass is defined first, followed by those used to organise the AddressSpace and then by the NodeClasses used to represent Objects.

The NodeClasses defined to represent Objects fall into three categories: those used to define instances, those used to define types for those instances and those used to define data types. Subclause 6.3 describes the rules for subtyping and 6.4 the rules for instantiation of the type definitions.

5.2 Base NodeClass

5.2.1 General

The OPC UA Address Space Model defines a Base NodeClass from which all other NodeClasses are derived. The derived NodeClasses represent the various components of the OPC UA Object Model (see 4.2). The Attributes of the Base NodeClass are specified in Table 2. There are no References specified for the Base NodeClass.

Table 2 – Base NodeClass

5.2.2 NodeId

Nodes are unambiguously identified using a constructed identifier called the NodeId. Some Servers may accept alternative NodeIds in addition to the canonical NodeId represented in this Attribute. A Server shall persist the NodeId of a Node, that is, it shall not generate new NodeIds when rebooting. The structure of the NodeId is defined in 8.2.

5.2.3 NodeClass

The NodeClass Attribute identifies the NodeClass of a Node. Its data type is defined in 8.30.

5.2.4 BrowseName

Nodes have a BrowseName Attribute that is used as a non-localised human-readable name when browsing the AddressSpace to create paths out of BrowseNames. The TranslateBrowsePathsToNodeIds Service defined in Part 4 can be used to follow a path constructed of BrowseNames.

A BrowseName should never be used to display the name of a Node. The DisplayName should be used instead for this purpose.

Unlike NodeIds, the BrowseName cannot be used to unambiguously identify a Node. Different Nodes may have the same BrowseName.

Subclause 8.3 defines the structure of the BrowseName. It contains a namespace and a string. The namespace is provided to make the BrowseName unique in some cases in the context of a Node (e.g. Properties of a Node) although not unique in the context of the Server. If different organizations define BrowseNames for Properties, the namespace of the BrowseName provided by the organization makes the BrowseName unique, although different organizations may use the same string having a slightly different meaning.

Servers may often choose to use the same namespace for the NodeId and the BrowseName. However, if they want to provide a standard Property, its BrowseName shall have the namespace of the standards body although the namespace of the NodeId reflects something else, for example the local Server.

It is recommended that standard bodies defining standard type definitions use their namespace for the NodeId of the TypeDefinitionNode as well as for the BrowseName of the TypeDefinitionNode.

The string-part of the BrowseName is case sensitive. That is, Clients shall consider them case sensitive. Servers are allowed to handle BrowseNames passed in Service requests as case insensitive. Examples are the TranslateBrowsePathsToNodeIds Service or Event filter.

5.2.5 DisplayName

The DisplayName Attribute contains the localised name of the Node. Clients should use this Attribute if they want to display the name of the Node to the user. They should not use the BrowseName for this purpose. The Server may maintain one or more localised representations for each DisplayName. Clients negotiate the locale to be returned when they open a session with the Server. Refer to Part 4 for a description of session establishment and locales. Subclause 8.5 defines the structure of the DisplayName. The string part of the DisplayName is restricted to 512 characters.

5.2.6 Description

The optional Description Attribute shall explain the meaning of the Node in a localised text using the same mechanisms for localisation as described for the DisplayName in 5.2.5.

5.2.7 WriteMask

The optional WriteMask Attribute exposes the possibilities of a client to write the Attributes of the Node. The WriteMask Attribute does not take any user access rights into account, that is, although an Attribute is writable this may be restricted to a certain user/user group. If the OPC UA Server does not have the ability to get the WriteMask information for a specific Attribute from the underlying system, it should state that it is writable. If a write operation is called on the Attribute, the Server should transfer this request and return the corresponding StatusCode if such a request is rejected. StatusCodes are defined in Part 4.

The WriteMask Attribute is a 32-bit unsigned integer with the structure defined in Table 3. If the bit is set to 0, it means the Attribute is not writable, if it is set to 1, it means it is writable. If a Node does not support a specific Attribute, the corresponding bit has to be set to 0.

Table 3 – Bit mask for WriteMask and UserWriteMask

5.2.8 UserWriteMask

The optional UserWriteMask Attribute exposes the possibilities of a client to write the Attributes of the Node taking user access rights into account. It uses the same bit mask as used in the WriteMask Attribute, defined in Table 3. The UserWriteMask Attribute can only further restrict the WriteMask Attribute, when it is set to not writable in the general case that applies for every user.

5.3 ReferenceType NodeClass

5.3.1 General

References are defined as instances of ReferenceType Nodes. ReferenceType Nodes are visible in the AddressSpace and are defined using the ReferenceType NodeClass as specified in Table 4. In contrast, a Reference is an inherent part of a Node and no NodeClass is used to represent References.

This standard defines a set of ReferenceTypes provided as an inherent part of the OPC UA Address Space Model. These ReferenceTypes are defined in Clause 7 and their representation in the AddressSpace is defined in Part 5. Servers may also define ReferenceTypes. In addition, Part 4 defines NodeManagement Services that allow Clients to add ReferenceTypes to the AddressSpace.

Table 4 – ReferenceType NodeClass

5.3.2 Attributes

The ReferenceType NodeClass inherits the base Attributes from the Base NodeClass defined in 5.2. The inherited BrowseName Attribute is used to specify the meaning of the ReferenceType as seen from the SourceNode. For example, the ReferenceType with the BrowseName “Contains” is used in References that specify that the SourceNode contains the TargetNode. The inherited DisplayName Attribute contains a translation of the BrowseName.

The BrowseName of a ReferenceType shall be unique in a Server. It is not allowed that two different ReferenceTypes have the same BrowseName.

The IsAbstract Attribute indicates if the ReferenceType is abstract. Abstract ReferenceTypes cannot be instantiated and are used only for organizational reasons, for example to specify some general semantics or constraints that its subtypes inherit.

The Symmetric Attribute is used to indicate whether or not the meaning of the ReferenceType is the same for both the SourceNode and TargetNode.

If a ReferenceType is symmetric, the InverseName Attribute shall be omitted. Examples of symmetric ReferenceTypes are “Connects To” and “Communicates With”. Both imply the same semantic coming from the SourceNode or the TargetNode. Therefore both directions are considered to be forward References.

If the ReferenceType is non-symmetric and not abstract, the InverseName Attribute shall be set. The InverseName Attribute specifies the meaning of the ReferenceType as seen from the TargetNode. Examples of non-symmetric ReferenceTypes include “Contains” and “Contained In”, and “Receives From” and “Sends To”.

References that use the InverseName, such as “Contained In” References, are referred to as inverse References.

Figure 8 provides examples of symmetric and non-symmetric References and the use of the BrowseName and the InverseName.

It might not always be possible for Servers to instantiate both forward and inverse References for non-symmetric ReferenceTypes as shown in Figure 8. When they do, the References are referred to as bidirectional. Although not required, it is recommended that all hierarchical References be instantiated as bidirectional to ensure browse connectivity. A bidirectional Reference is modelled as two separate References. As an example of a unidirectional Reference, it is often the case that a subscriber knows its publisher, but its publisher does not know its subscribers. The subscriber would have a “Subscribes To” Reference to the publisher, without the publisher having the corresponding “Publishes To” inverse References to its subscribers.

The DisplayName and the InverseName are the only standardised places to indicate the semantic of a ReferenceType. There may be more complex semantics associated with a ReferenceType than can be expressed in those Attributes (e.g. the semantic of HasSubtype). This standard does not specify how this semantic should be exposed. However, the Description Attribute can be used for this purpose. This standard provides a semantic for the ReferenceTypes specified in Clause 7.

A ReferenceType can have constraints restricting its use. For example, it can specify that starting from Node A and only following References of this ReferenceType or one of its subtypes, it shall never be able to return to A, that is, a “No Loop” constraint.

This standard does not specify how those constraints could or should be made available in the AddressSpace. Nevertheless, for the standard ReferenceTypes, some constraints are specified in Clause 7. This standard does not restrict the kind of constraints valid for a ReferenceType. It can, for example, also affect an ObjectType. The restriction that a ReferenceType can only be used by relating Nodes of some NodeClasses with a defined cardinality is a special constraint of a ReferenceType.

5.3.3 References

5.3.3.1 General

HasSubtype References and HasProperty References are the only ReferenceTypes that may be used with ReferenceType Nodes as SourceNode. ReferenceType Nodes shall not be the SourceNode of other types of References.

5.3.3.2 HasProperty References

HasProperty References are used to identify the Properties of a ReferenceType and shall only refer to Nodes of the Variable NodeClass.

The Property NodeVersion is used to indicate the version of the ReferenceType.

There are no additional Properties defined for ReferenceTypes in this standard. Additional parts this series of standards may define additional Properties for ReferenceTypes.

5.3.3.3 HasSubtype References

HasSubtype References are used to define subtypes of ReferenceTypes. It is not required to provide the HasSubtype Reference for the supertype, but it is required that the subtype provides the inverse Reference to its supertype. The following rules for subtyping apply.

a) The semantic of a ReferenceType (e.g. “spans a hierarchy”) is inherited to its subtypes and can be refined there (e.g. “spans a special hierarchy”). The DisplayName, and also the InverseName for non-symmetric ReferenceTypes, reflect the specialization. b) If a ReferenceType specifies some constraints (e.g. “allow no loops”) this is inherited and can only be refined (e.g. inheriting “no loops” could be refined as “shall be a tree – only one parent”) but not lowered (e.g. “allow loops”). c) The constraints concerning which NodeClasses can be referenced are also inherited and can only be further restricted. That is, if a ReferenceType “A” is not allowed to relate an Object with an ObjectType, this is also true for its subtypes. d) A ReferenceType shall have exactly one supertype, except for the References ReferenceType defined in 7.2 as the root type of the ReferenceType hierarchy. The ReferenceType hierarchy does not support multiple inheritances.

5.4 View NodeClass

Underlying systems are often large and Clients often have an interest in only a specific subset of the data. They do not need, or want, to be burdened with viewing Nodes in the AddressSpace for which they have no interest.

To address this problem, this standard defines the concept of a View. Each View defines a subset of the Nodes in the AddressSpace. The entire AddressSpace is the default View. Each Node in a View may contain only a subset of its References, as defined by the creator of the View. The View Node acts as the root for the Nodes in the View. Views are defined using the View NodeClass, which is specified in Table 5.

All Nodes contained in a View shall be accessible starting from the View Node when browsing in the context of the View. It is not expected that all containing Nodes can be browsed directly from the View Node but rather browsed from other Nodes contained in the View.

A View Node may not only be used as additional entry point into the AddressSpace but as a construct to organize the AddressSpace and thus as the only entry point into a subset of the AddressSpace. Therefore Clients shall not ignore View Nodes when exposing the AddressSpace. Simple Clients that do not deal with Views for filtering purposes can, for example, handle a View Node like an Object of type FolderType (see 5.5.3).

Table 5 – View NodeClass

The View NodeClass inherits the base Attributes from the Base NodeClass defined in 5.2. It also defines two additional Attributes.

The mandatory ContainsNoLoops Attribute is set to false if the Server is not able to identify if the View contains loops or not.

The mandatory EventNotifier Attribute identifies if the View can be used to subscribe to Events that either occur in the content of the View or as ModelChangeEvents (see 9.32) of the content of the View or to read / write the history of the Events. A View that supports Events shall provide all Events that occur in any Object used as EventNotifier that is part of the content of the View. In addition, it shall provide all ModelChangeEvents that occur in the context of the View.

To avoid recursion, i.e. getting all Events of the Server, the Server Object defined in Part 5 shall never be part of any View since it provides all Events of the Server.

Views are defined by the Server. The browsing and querying Services defined in Part 4 expect the NodeId of a View Node to provide these Services in the context of the View.

HasProperty References are used to identify the Properties of a View. The Property NodeVersion is used to indicate the version of the View Node. The ViewVersion Property indicates the version of the content of the View. In contrast to the NodeVersion, the ViewVersion Property is updated even if Nodes not directly referenced by the View Node are added to or deleted from the View. This Property is optional because it might not be possible for Servers to detect changes in the View contents. Servers may also generate a ModelChangeEvent, described in 9.32, if Nodes are added to or deleted from the View. There are no additional Properties defined for Views in this document. Additional parts of this series of standards may define additional Properties for Views.

Views can be the SourceNode of any hierarchical Reference. They shall not be the SourceNode of any non-hierarchical Reference.

5.5 Objects

5.5.1 Object NodeClass

Objects are used to represent systems, system components, real-world objects and software objects. Objects are defined using the Object NodeClass, specified in Table 6.

Table 6 – Object NodeClass

The Object NodeClass inherits the base Attributes from the Base NodeClass defined in 5.2.

The mandatory EventNotifier Attribute identifies whether the Object can be used to subscribe to Events or to read and write the history of the Events.

The Object NodeClass uses the HasComponent Reference to define the DataVariables, Objects and Methods of an Object.

It uses the HasProperty Reference to define the Properties of an Object. The Property NodeVersion is used to indicate the version of the Object. The Property Icon provides an icon of the Object. The Property NamingRule defines the NamingRule of a ModellingRule and shall only be applied to Objects of type ModellingRuleType. There are no additional Properties defined for Objects in this document. Additional parts of this series of standards may define additional Properties for Objects.

To specify its ModellingRule, an Object can use at most one HasModellingRule Reference pointing to a ModellingRule Object. ModellingRules are defined in 6.4.4.

HasNotifier and HasEventSource References are used to provide information about eventing and can only be applied to Objects used as event notifiers. Details are defined in 7.17 and 7.19.

The HasTypeDefinition Reference points to the ObjectType used as type definition of the Object.

Objects may use any additional References to define relationships to other Nodes. No restrictions are placed on the types of References used or on the NodeClasses of the Nodes that may be referenced. However, restrictions may be defined by the ReferenceType excluding its use for Objects. Standard ReferenceTypes are described in Clause 7.

If the Object is used as an InstanceDeclaration (see 4.5) then all Nodes referenced with hierarchical References in a forward direction shall have unique BrowseNames in the context of this Object.

If the Object is created based on an InstanceDeclaration then it shall have the same BrowseName as its InstanceDeclaration.

5.5.2 ObjectType NodeClass

ObjectTypes provide definitions for Objects. ObjectTypes are defined using the ObjectType NodeClass, which is specified in Table 7.

Table 7 – ObjectType NodeClass

The ObjectType NodeClass inherits the base Attributes from the Base NodeClass defined in 5.2. The additional IsAbstract Attribute indicates if the ObjectType is abstract or not.

The ObjectType NodeClass uses the HasComponent References to define the DataVariables, Objects, and Methods for it.

The HasProperty Reference is used to identify the Properties. The Property NodeVersion is used to indicate the version of the ObjectType. The Property Icon provides an icon of the ObjectType. There are no additional Properties defined for ObjectTypes in this document. Additional parts of this series of standards may define additional Properties for ObjectTypes.

HasSubtype References are used to subtype ObjectTypes. ObjectType subtypes inherit the general semantics from the parent type. The general rules for subtyping apply as defined in Clause 6. It is not required to provide the HasSubtype Reference for the supertype, but it is required that the subtype provides the inverse Reference to its supertype.

GeneratesEvent References identify the type of Events that instances of the ObjectType may generate. These Objects may be the source of an Event of the specified type or one of its subtypes. Servers should make GeneratesEvent References bidirectional References. However, it is allowed to be unidirectional when the Server is not able to expose the inverse direction pointing from the EventType to each ObjectType supporting the EventType. Note that the EventNotifier Attribute of an Object and the GeneratesEvent References of its ObjectType are completely unrelated. Objects that can generate Events might not be used as Objects to which Clients subscribe to get the corresponding Event notifications.

GeneratesEvent References are optional, i.e. Objects may generate Events of an EventType that is not exposed by its ObjectType.

ObjectTypes may use any additional References to define relationships to other Nodes. No restrictions are placed on the types of References used or on the NodeClasses of the Nodes that may be referenced. However, restrictions may be defined by the ReferenceType excluding its use for ObjectTypes. Standard ReferenceTypes are described in Clause 7.

All Nodes referenced with hierarchical References shall have unique BrowseNames in the context of an ObjectType (see 4.5).

5.5.3 Standard ObjectType FolderType

The ObjectType FolderType is formally defined in Part 5. Its purpose is to provide Objects that have no other semantic than organizing of the AddressSpace. A special ReferenceType is introduced for those Folder Objects, the Organizes ReferenceType. The SourceNode of such a Reference should always be a View or an Object of the ObjectType FolderType; the TargetNode can be of any NodeClass. Organizes References can be used in any combination with HasChild References (HasComponent, HasProperty, etc.; see 7.5) and do not prevent loops. Thus, they can be used to span multiple hierarchies.

5.5.4 Client-side creation of Objects of an ObjectType

Objects are always based on an ObjectType, i.e. they have a HasTypeDefinition Reference pointing to its ObjectType.

Clients can create Objects using the AddNodes Service defined in Part 4. The Service requires specifying the TypeDefinitionNode of the Object. An Object created by the AddNodes Service contains all components defined by its ObjectType dependent on the ModellingRules specified for the components. However, the Server may add additional components and References to the Object and its components that are not defined by the ObjectType. This behaviour is Server dependent. The ObjectType only specifies the minimum set of components that shall exist for each Object of an ObjectType.

In addition to the AddNodes Service ObjectTypes may have a special Method with the BrowseName “Create”. This Method is used to create an Object of this ObjectType. This Method may be useful for the creation of Objects where the semantic of the creation should differ from the default behaviour expected in the context of the AddNodes Service. For example, the values should directly differ from the default values or additional Objects should be added, etc. The input and output arguments of this Method depend on the ObjectType; the only commonality is the BrowseName identifying that this Method will create an Object based on the ObjectType. Servers should not provide a Method on an ObjectType with the BrowseName “Create” for any other purpose than creating Objects of the ObjectType.

5.6 Variables

5.6.1 General

Two types of Variables are defined, Properties and DataVariables. Although they differ in the way they are used as described in 4.4 and have different constraints described in the remainder of 5.6 they use the same NodeClass described in 5.6.2. The constraints of Properties based on this NodeClass are defined in 5.6.3, the constraints of DataVariables in 5.6.4.

5.6.2 Variable NodeClass

Variables are used to represent values which may be simple or complex. Variables are defined by VariableTypes, as specified in 5.6.5.

Variables are always defined as Properties or DataVariables of other Nodes in the AddressSpace. They are never defined by themselves. A Variable is always part of at least one other Node, but may be related to any number of other Nodes. Variables are defined using the Variable NodeClass, specified in Table 8.

Table 8 – Variable NodeClass

The UserAccessLevel is an 8-bit unsigned integer with the structure defined in the following table:

The Variable NodeClass inherits the base Attributes from the Base NodeClass defined in 5.2.

The Variable NodeClass also defines a set of Attributes that describe the Variable’s Runtime value. The Value Attribute represents the Variable value. The DataType, ValueRank and ArrayDimensions Attributes provide the capability to describe simple and complex values.

The AccessLevel Attribute indicates the accessibility of the Value of a Variable not taking user access rights into account. If the OPC UA Server does not have the ability to get the AccessLevel information from the underlying system then it should state that it is readable and writable. If a read or write operation is called on the Variable then the Server should transfer this request and return the corresponding StatusCode even if such a request is rejected. StatusCodes are defined in Part 4.

The SemanticChange bit of the AccessLevel Attribute shall be set when the Property describes the semantic of the Node that owns the Property and changes of the Property value will generate SemanticChangeEvents. For example, a Property describing the engineering unit of a DataVariable will have the bit set, whereas a Property containing an Icon of the DataVariable will not. This behaviour is exactly the same as described by the SemanticsChanged bit of the StatusCode defined in Part 4. However, if subscribing to a Variable then one should look at the StatusCode to identify if the semantic has changed in order to receive this information before processing the value of the Variable.

The UserAccessLevel Attribute indicates the accessibility of the Value of a Variable taking user access rights into account. If the OPC UA Server does not have the ability to get any user access rights related information from the underlying system then it should use the same bit mask as used in the AccessLevel Attribute. The UserAccessLevel Attribute can restrict the accessibility indicated by the AccessLevel Attribute, but not exceed it.

The MinimumSamplingInterval Attribute specifies how fast the Server can reasonably sample the value for changes. The accuracy of this value (the ability of the Server to attain “best case” performance) can be greatly affected by system load and other factors.

The Historizing Attribute indicates whether the Server is actively collecting data for the history of the Variable. See Part 11 for details on historizing Variables.

Clients may read or write Variable values, or monitor them for value changes, as specified in Part 4. Part 8 defines additional rules when using the Services for automation data.

To specify its ModellingRule, a Variable can use at most one HasModellingRule Reference pointing to a ModellingRule Object. ModellingRules are defined in 6.4.4.

If the Variable is created based on an InstanceDeclaration (see 4.5) it shall have the same BrowseName as its InstanceDeclaration.

The other References are described separately for Properties and DataVariables in the remainder of 5.6

5.6.3 Properties

Properties are used to define the characteristics of Nodes. Properties are defined using the Variable NodeClass, specified in Table 8. However, they restrict their use.

Properties are the leaf of any hierarchy; therefore they shall not be the SourceNode of any hierarchical References. This includes the HasComponent or HasProperty Reference, that is, Properties do not contain Properties and cannot expose their complex structure. However, they may be the SourceNode of any non-hierarchical References.

The HasTypeDefinition Reference points to the VariableType of the Property. Since Properties are uniquely identified by their BrowseName, all Properties shall point to the PropertyType defined in Part 5.

Properties shall always be defined in the context of another Node and shall be the TargetNode of at least one HasProperty Reference. To distinguish them from DataVariables, they shall not be the TargetNode of any HasComponent Reference. Thus, a HasProperty Reference pointing to a Variable Node defines this Node as a Property.

The BrowseName of a Property is always unique in the context of a Node. It is not permitted for a Node to refer to two Variables using HasProperty References having the same BrowseName.

5.6.4 DataVariable

DataVariables represent the content of an Object. DataVariables are defined using the Variable NodeClass, specified in Table 8.

DataVariables identify their Properties using HasProperty References. Complex DataVariables use HasComponent References to expose their component DataVariables.

The Property NodeVersion indicates the version of the DataVariable. The Property LocalTime indicates the difference between the SourceTimestamp of the value and the standard time at the location in which the value was obtained. The Property DataTypeVersion is used only for DataTypeDictionaries and DataTypeDescriptions as defined in 5.8. The Standard Property DictionaryFragment is used only for DataTypeDescriptions as defined in 5.8. The Property AllowNulls indicates if null values are allowed for the Value Attribute. The Property ValueAsText provides a localized text representation for enumeration values. The Property MaxStringLength indicates the maximum allowed length of a string value, the Property MaxByteStringLength the maximum allowed length of a byte string value and the Property MaxArrayLength the maximum allowed array length of the value. The Property EngineeringUnits indicates the engineering units of the value. There are no additional Properties defined for DataVariables in this part of this document. Additional parts of this series of standards may define additional Properties for DataVariables. Part 8 defines a set of Properties that can be used for DataVariables.

DataVariables may use additional References to define relationships to other Nodes. No restrictions are placed on the types of References used or on the NodeClasses of the Nodes that may be referenced. However, restrictions may be defined by the ReferenceType excluding its use for DataVariables. Standard ReferenceTypes are described in Clause 7.

A DataVariable is intended to be defined in the context of an Object. However, complex DataVariables may expose other DataVariables, and ObjectTypes and complex VariableTypes may also contain DataVariables. Therefore each DataVariable shall be the TargetNode of at least one HasComponent Reference coming from an Object, an ObjectType, a DataVariable or a VariableType. DataVariables shall not be the TargetNode of any HasProperty References. Therefore, a HasComponent Reference pointing to a Variable Node identifies it as a DataVariable.

The HasTypeDefinition Reference points to the VariableType used as type definition of the DataVariable.

If the DataVariable is used as InstanceDeclaration (see 4.5) all Nodes referenced with hierarchical References in the forward direction shall have unique BrowseNames in the context of this DataVariable.

5.6.5 VariableType NodeClass

VariableTypes are used to provide type definitions for Variables. VariableTypes are defined using the VariableType NodeClass, as specified in Table 9.

Table 9 – VariableType NodeClass

A value of 0 for an individual dimension indicates that the dimension has a variable length.
For example, if a VariableType is defined by the following C array:
Int32 myArray[346];
then this VariableType’s DataType would point to an Int32, the VariableType’s ValueRank has the value 1 and the ArrayDimensions is an array with one entry having the value 346.| |IsAbstract|M|Boolean|A boolean Attribute with the following values:

• **TRUE**: it is an abstract VariableType, i.e. no Variable of this type shall exist, only of its subtypes.
• FALSE it is not an abstract VariableType, i.e. Variables of this type can exist.

The VariableType NodeClass inherits the base Attributes from the Base NodeClass defined in 5.2. The VariableType NodeClass also defines a set of Attributes that describe the default or initial value of its instance Variables. The Value Attribute represents the default value. The DataType, ValueRank and ArrayDimensions Attributes provide the capability to describe simple and complex values. The IsAbstract Attribute defines if the type can be directly instantiated.

The VariableType NodeClass uses HasProperty References to define the Properties and HasComponent References to define DataVariables. Whether they are instantiated depends on the ModellingRules defined in 6.4.4.

The Property NodeVersion indicates the version of the VariableType. There are no additional Properties defined for VariableTypes in this document. Additional parts of this series of standards may define additional Properties for VariableTypes. Part 8 defines a set of Properties that can be used for VariableTypes.

HasSubtype References are used to subtype VariableTypes. VariableType subtypes inherit the general semantics from the parent type. The general rules for subtyping are defined in Clause 6. It is not required to provide the HasSubtype Reference for the supertype, but it is required that the subtype provides the inverse Reference to its supertype.

GeneratesEvent References identify that Variables of the VariableType may be the source of an Event of the specified EventType or one of its subtypes. Servers should make GeneratesEvent References bidirectional References. However, it is allowed to be unidirectional when the Server is not able to expose the inverse direction pointing from the EventType to each VariableType supporting the EventType.

GeneratesEvent References are optional, i.e. Variables may generate Events of an EventType that is not exposed by its VariableType.

VariableTypes may use any additional References to define relationships to other Nodes. No restrictions are placed on the types of References used or on the NodeClasses of the Nodes that may be referenced. However, restrictions may be defined by the ReferenceType excluding its use for VariableTypes. Standard ReferenceTypes are described in Clause 7.

All Nodes referenced with hierarchical References shall have unique BrowseNames in the context of the VariableType (see 4.5).

5.6.6 Client-side creation of Variables of an VariableType

Variables are always based on a VariableType, i.e. they have a HasTypeDefinition Reference pointing to its VariableType.

Clients can create Variables using the AddNodes Service defined in Part 4. The Service requires specifying the TypeDefinitionNode of the Variable. A Variable created by the AddNodes Service contains all components defined by its VariableType dependent on the ModellingRules specified for the components. However, the Server may add additional components and References to the Variable and its components that are not defined by the VariableType. This behaviour is Server dependent. The VariableType only specifies the minimum set of components that shall exist for each Variable of a VariableType.

5.7 Method NodeClass

Methods define callable functions. Methods are invoked using the Call Service defined in Part 4. Method invocations are not represented in the AddressSpace. Method invocations always run to completion and always return responses when complete. Methods are defined using the Method NodeClass, specified in Table 10.

Table 10 – Method NodeClass

The Method NodeClass inherits the base Attributes from the Base NodeClass defined in 5.2. The Method NodeClass defines no additional Attributes.

The Executable Attribute indicates whether the Method is executable, not taking user access rights into account. If the OPC UA Server cannot get the Executable information from the underlying system, it should state that it is executable. If a Method is called then the Server should transfer this request and return the corresponding StatusCode even if such a request is rejected. StatusCodes are defined in Part 4.

The UserExecutable Attribute indicates whether the Method is executable, taking user access rights into account. If the OPC UA Server cannot get any user rights related information from the underlying system, it should use the same value as used in the Executable Attribute. The UserExecutable Attribute can be set to “False”, even if the Executable Attribute is set to “True”, but it shall be set to “False” if the Executable Attribute is set to “False”.

Properties may be defined for Methods using HasProperty References. The Properties InputArguments and OutputArguments specify the input arguments and output arguments of the Method. Both contain an array of the DataType Argument as specified in 8.6. An empty array or a Property that is not provided indicates that there are no input arguments or output arguments for the Method. The Property NodeVersion indicates the version of the Method. There are no additional Properties defined for Methods in this document. Additional parts of this series of standards may define additional Properties for Methods.

To specify its ModellingRule, a Method can use at most one HasModellingRule Reference pointing to a ModellingRule Object. ModellingRules are defined in 6.4.4.

GeneratesEvent References identify that Methods may generate an Event of the specified EventType or one of its subtypes for every call of the Method. A Server may generate one Event for each referenced EventType when a Method is successfully called.

AlwaysGeneratesEvent References identify that Methods will generate an Event of the specified EventType or one of its subtypes for every call of the Method. A Server shall always generate one Event for each referenced EventType when a Method is successfully called.

Servers should make GeneratesEvent References bidirectional References. However, it is allowed to be unidirectional when the Server is not able to expose the inverse direction pointing from the EventType to each Method generating the EventType.

GeneratesEvent References are optional, i.e. the call of a Method may produce Events of an EventType that is not referenced with a GeneratesEvent Reference by the Method.

Methods may use additional References to define relationships to other Nodes. No restrictions are placed on the types of References used or on the NodeClasses of the Nodes that may be referenced. However, restrictions may be defined by the ReferenceType excluding its use for Methods. Standard ReferenceTypes are described in Clause 7.

A Method shall always be the TargetNode of at least one HasComponent Reference. The SourceNode of these HasComponent References shall be an Object or an ObjectType. If a Method is called then the NodeId of one of those Nodes shall be put into the Call Service defined in Part 4 as parameter to detect the context of the Method operation.

If the Method is used as InstanceDeclaration (see 4.5) all Nodes referenced with hierarchical References in forward direction shall have unique BrowseNames in the context of this Method.

5.8 DataTypes

5.8.1 DataType Model

The DataType Model is used to define simple and structured data types. Data types are used to describe the structure of the Value Attribute of Variables and their VariableTypes. Therefore each Variable and VariableType is pointing with its DataType Attribute to a Node of the DataType NodeClass as shown in Figure 9.

Figure 9 - Variables, VariableTypes and their DataTypes

In many cases, the NodeId of the DataType Node – the DataTypeId – will be well-known to Clients and Servers. Clause 8 defines DataTypes and Part 6 defines their DataTypeIds. In addition, other organizations may define DataTypes that are well-known in the industry. Well- known DataTypeIds provide for commonality across OPC UA Servers and allow Clients to interpret values without having to read the type description from the Server. Therefore, Servers may use well-known DataTypeIds without representing the corresponding DataType Nodes in their AddressSpaces.

In other cases, DataTypes and their corresponding DataTypeIds may be vendor-defined. Servers should attempt to expose the DataType Nodes and the information about the structure of those DataTypes for Clients to read, although this information might not always be available to the Server.

Figure 10 illustrates the Nodes used in the AddressSpace to describe the structure of a DataType. The DataType points to an Object of type DataTypeEncodingType. Each DataType can have several DataTypeEncoding, for example “Default”, “UA Binary” and “XML” encoding. Services in Part 4 allow Clients to request an encoding or choosing the “Default” encoding.

Each DataTypeEncoding is used by exactly one DataType, that is, it is not permitted for two DataTypes to point to the same DataTypeEncoding. The DataTypeEncoding Object points to exactly one Variable of type DataTypeDescriptionType. The DataTypeDescription Variable belongs to a DataTypeDictionary Variable.

Figure 10 - DataType Model

Since the NodeId of the DataTypeEncoding will be used in some Mappings to identify the DataType and its encoding as defined in Part 6, those NodeIds may also be well-known for well- known DataTypeIds.

The DataTypeDictionary describes a set of DataTypes in sufficient detail to allow Clients to parse/interpret Variable Values that they receive and to construct Values that they send. The DataTypeDictionary is represented as a Variable of type DataTypeDictionaryType in the AddressSpace, the description about the DataTypes is contained in its Value Attribute. All containing DataTypes exposed in the AddressSpace are represented as Variables of type DataTypeDescriptionType. The Value of one of these Variables identifies the description of a DataType in the Value Attribute of the DataTypeDictionary.

The DataType of a DataTypeDictionary Variable is always a ByteString. The format and conventions for defining DataTypes in this ByteString are defined by DataTypeSystems. DataTypeSystems are identified by NodeIds. They are represented in the AddressSpace as Objects of the ObjectType DataTypeSystemType. Each Variable representing a DataTypeDictionary references a DataTypeSystem Object to identify their DataTypeSystem.

A client shall recognise the DataTypeSystem to parse any of the type description information. OPC UA Clients that do not recognise a DataTypeSystem will not be able to interpret its type descriptions, and consequently, the values described by them. In these cases, Clients interpret these values as opaque ByteStrings.

OPC Binary and W3C XML Schema are examples of DataTypeSystems. The OPC Binary DataTypeSystem is defined in Annex C. OPC Binary uses XML to describe binary data values. W3C XML Schema is specified in XML Schema Part 1 and XML Schema Part 2

5.8.2 Encoding Rules for different kinds of DataTypes

Different kinds of DataTypes are handled differently regarding their encoding and according to whether this encoding is represented in the AddressSpace.

Built-in DataTypes are a fixed set of DataTypes (see Part 6 for a complete list of Built-in DataTypes). They have no encodings visible in the AddressSpace since the encoding should be known to all OPC UA products. Examples of Built-in DataTypes are Int32 (see 8.26) and Double (see 8.12).

Simple DataTypes are subtypes of the Built-in DataTypes. They are handled on the wire like the Built-in DataType, i.e. they cannot be distinguished on the wire from their Built-in supertypes. Since they are handled like Built-in DataTypes regarding the encoding they cannot have encodings defined in the AddressSpace. Clients can read the DataType Attribute of a Variable or VariableType to identify the Simple DataType of the Value Attribute. An example of a Simple DataType is Duration. It is handled on the wire as a Double but the Client can read the DataType Attribute and thus interpret the value as defined by Duration (see 8.13).

Structured DataTypes are DataTypes that represent structured data and are not defined as Built-in DataTypes. Structured DataTypes inherit directly or indirectly from the DataType Structure defined in 8.33. Structured DataTypes may have several encodings and the encodings are exposed in the AddressSpace. How the encoding of Structured DataTypes is handled on the wire is defined in Part 6. The encoding of the Structured DataType is transmitted with each value, thus Clients are aware of the DataType without reading the DataType Attribute. The encoding has to be transmitted so the Client is able to interpret the data. An example of a Structured DataType is Argument (see 8.6).

Enumeration DataTypes are DataTypes that represent discrete sets of named values. Enumerations are always encoded as Int32 on the wire as defined in Part 6. Enumeration DataTypes inherit directly or indirectly from the DataType Enumeration defined in 8.14. Enumerations have no encodings exposed in the AddressSpace. To expose the human- readable representation of an enumerated value the DataType Node may have the EnumStrings Property that contains an array of LocalizedText. The Integer representation of the enumeration value points to a position within that array. EnumValues Property can be used instead of the EnumStrings to support integer representation of enumerations that are not zero-based or have gaps. It contains an array of a Structured DataType containing the integer representation as well as the human-readable representation. An example of an enumeration DataType containing a sparse list of Integers is NodeClass which is defined in 8.30.

In addition to the DataTypes described above, abstract DataTypes are also supported, which do not have any encodings and cannot be exchanged on the wire. Variables and VariableTypes use abstract DataTypes to indicate that their Value may be any one of the subtypes of the abstract DataType. An example of an abstract DataType is Integer which is defined in 8.24.

5.8.3 DataType NodeClass

The DataType NodeClass describes the syntax of a Variable Value. The DataTypes may be simple or complex, depending on the DataTypeSystem. DataTypes are defined using the DataType NodeClass, as specified in Table 11.

Table 11 – DataType NodeClass |Name|Use|DataType|Description| |-----|-----|-----|------| |Attributes|||| |Base NodeClass Attributes|M|--|Inherited from the Base NodeClass. See 5.2| |IsAbstract|M|Boolean|A boolean Attribute with the following values:

• **TRUE** it is an abstract DataType.
• **FALSE** it is not an abstract DataType.

The DataType NodeClass inherits the base Attributes from the Base NodeClass defined in 5.2. The IsAbstract Attribute specifies if the DataType is abstract or not. Abstract DataTypes can be used in the AddressSpace, i.e. Variables and VariableTypes can point with their DataType Attribute to an abstract DataType. However, concrete values can never be of an abstract DataType and shall always be of a concrete subtype of the abstract DataType.

HasProperty References are used to identify the Properties of a DataType. The Property NodeVersion is used to indicate the version of the DataType. This Version is not affected by the DataTypeVersion Property of DataTypeDictionaries and DataTypeDescriptions. The Property EnumStrings contains human-readable representations of enumeration values and is only applied to Enumeration DataTypes. Instead of the EnumStrings Property an Enumeration DataType can also use the EnumValues Property to represent Enumerations with integer values that are not zero-based or containing gaps. There are no additional Properties defined for DataTypes in this standard. Additional parts of this series of standards may define additional Properties for DataTypes.

HasSubtype References may be used to expose a data type hierarchy in the AddressSpace. This hierarchy shall reflect the hierarchy specified in the DataTypeDictionary. The semantic of subtyping depends on the DataTypeSystem. Servers need not provide HasSubtype References, even if their DataTypes span a type hierarchy. Clients should not make any assumptions about any other semantic with that information than provided by the DataTypeDictionary. For example, it might not be possible to cast a value of one data type to its base data type. Some restrictions apply for subtyping enumeration DataTypes as defined in 8.14.

HasEncoding References point from the DataType to its DataTypeEncodings. Following such a Reference, the client can browse to the DataTypeDictionary describing the structure of the DataType for the used encoding. Each concrete Structured DataType can point to many DataTypeEncodings, but each DataTypeEncoding shall belong to one DataType, that is, it is not permitted for two DataType Nodes to point to the same DataTypeEncoding Object using HasEncoding References.

An abstract DataType is not the SourceNode of a HasEncoding Reference. The DataTypeEncoding of an abstract DataType is provided by its concrete subtypes.

DataType Nodes shall not be the SourceNode of other types of References. However, they may be the TargetNode of other References.

5.8.4 DataTypeDictionary, DataTypeDescription, DataTypeEncoding and DataTypeSystem

A DataTypeDictionary is an entity that contains a set of type descriptions, such as an XML schema. DataTypeDictionaries are defined as Variables of the VariableType DataTypeDictionaryType.

A DataTypeSystem specifies the format and conventions for defining DataTypes in DataTypeDictionaries. DataTypeSystems are defined as Objects of the ObjectType DataTypeSystemType.

The ReferenceType used to relate Objects of the ObjectType DataTypeSystemType to Variables of the VariableType DataTypeDictionaryType is the HasComponent ReferenceType. Thus, the Variable is always the TargetNode of a HasComponent Reference; this is a requirement for Variables. However, for DataTypeDictionaries the Server shall always provide the inverse Reference, since it is necessary to know the DataTypeSystem when processing the DataTypeDictionary.

Changes may be a result of a change to a type description, but it is more likely that dictionary changes are a result of the addition or deletion of type descriptions. This includes changes made while the Server is offline so that the new version is available when the Server restarts. Clients may subscribe to the DataTypeVersion Property to determine if the DataTypeDictionary has changed since it was last read.

The Server may, but is not required to, make the DataTypeDictionary contents available to Clients through the Value Attribute. Clients should assume that DataTypeDictionary contents are relatively large and that they will encounter performance problems if they automatically read the DataTypeDictionary contents each time they encounter an instance of a specific DataType. The client should use the DataTypeVersion Property to determine whether the locally cached copy is still valid. If the client detects a change to the DataTypeVersion, then it shall re-read the DataTypeDictionary. This implies that the DataTypeVersion shall be updated by a Server even after restart since Clients may persistently store the locally cached copy.

The Value Attribute of the DataTypeDictionary containing the type descriptions is a ByteString whose formatting is defined by the DataTypeSystem. For the “XML Schema” DataTypeSystem, the ByteString contains a valid XML Schema document. For the “OPC Binary” DataTypeSystem, the ByteString contains a string that is a valid XML document. The Server shall ensure that any change to the contents of the ByteString is matched with a corresponding change to the DataTypeVersion Property. In other words, the client may safely use a cached copy of the

DataTypeDictionary, as long as the DataTypeVersion remains the same.

DataTypeDictionaries are complex Variables which expose their DataTypeDescriptions as Variables using HasComponent References. A DataTypeDescription provides the information necessary to find the formal description of a DataType within the DataTypeDictionary. The Value of a DataTypeDescription depends on the DataTypeSystem of the DataTypeDictionary. When using “OPC Binary” dictionaries the Value shall be the name of the TypeDescription. When using “XML Schema” dictionaries the Value shall be an Xpath expression (see XPATH) which points to an XML element in the schema document.

Like DataTypeDictionaries each DataTypeDescription provides the Property DataTypeVersion indicating whether the type description of the DataType has changed. Changes to the DataTypeVersion may impact the operation of Subscriptions. If the DataTypeVersion changes for a Variable that is being monitored for a Subscription and that uses this DataTypeDescription, then the next data change Notification sent for the Variable will contain a status that indicates the change in the DataTypeDescription.

DataTypeEncoding Objects of the DataTypes reference their DataTypeDescriptions of the DataTypeDictionaries using HasDescription References. Servers shall provide the inverse References that relate the DataTypeDescriptions back to the DataTypeEncoding Objects. If a DataType Node is exposed in the AddressSpace, it shall provide its DataTypeEncodings and if a DataTypeDictionary is exposed then it should expose all of its DataTypeDescriptions. Both of these References shall be bi-directional.

The VariableTypes DataTypeDictionaryType and DataTypeDescriptionType and the ObjectTypes DataTypeSystemType and DataTypeEncodingType are formally defined in Part 5.

Figure 11 provides an example how DataTypes are modelled in the AddressSpace.

Figure 11 – Example of DataType Modelling

In some scenarios an OPC UA Server may have resource limitations which make it impractical to expose large DataTypeDictionaries. In these scenarios the Server may be able to provide access to descriptions for individual DataTypes even if the entire dictionary cannot be read. For this reason, this standard defines a Property for the DataTypeDescription called DictionaryFragment (see 5.6.2). This Property is a ByteString that contains a subset of the DataTypeDictionary which describes the format of the DataType associated with the DataTypeDescription. Thus, the Server splits the large DataTypeDictionary into several small parts and Clients can access without affecting the overall system performance.

However, Servers should provide the whole DataTypeDictionary at once if this is possible. It is typically more efficient to read the whole DataTypeDictionary at once instead of reading individual parts.

5.9 Summary of Attributes of the NodeClasses

Table 12 summarises all Attributes defined in this document and points out which NodeClasses use them either in an optional (O) or mandatory (M) way.

Table 12 – Overview of Attributes